BackgroundIt has previously been demonstrated that an extensive upper limb neurological examination of individual muscle function, sensation in homonymous innervated territories, and nerve trunk allodynia is reliable and that the outcome reflects symptoms. Since this approach may appear complicated and time consuming, this study deals with the value of an examination limited to manual testing of only six muscles.MethodsTwo examiners blinded to symptom status performed manual muscle testing of six muscles in 82 upper limbs with or without pain, weakness, and/or numbness/tingling. The six muscles represent three antagonist pairs (pectoralis major/posterior deltoid, biceps/triceps, and radial flexor of wrist/short radial extensor of wrist). The inter-rater reliability of detecting muscular weaknesses and the relation of weakness to the mentioned symptoms were analysed by kappa-statistics.ResultsThe two examiners recognized weaknesses in 48 and 55 limbs, respectively, with moderate agreement (median kappa = 0.58). Out of these, 35 and 32 limbs, respectively, were symptomatic. There was good correlation between findings and symptoms for one examiner (kappa = 0.61) and fair correlation for the other one (kappa = 0.33). Both reached high sensitivity (0.92, 0.84) but less satisfactory specificity (0.70, 0.50). Weaknesses agreed upon by the two examiners correlated moderately with symptoms (kappa = 0.57).ConclusionsWeakness in one or more muscles was present in almost all symptomatic limbs but in many non-symptomatic limbs as well. Manual testing of six muscles may represent a useful screening approach to upper limb neuropathic conditions, but a confirmative diagnosis requires further assessment.